ClimaX: A foundation model for weather and climate

Source: arXiv

 

In this publication, researchers Tung NguyenJohannes BrandstetterAshish KapoorJayesh K. GuptaAditya Grover have developed a weather modeling program called ClimaXClimaX is based on a deep neural network architecture. 

 

Using Deep Learning, ClimaX algorithms are trained to predict atmospheric variables with heterogeneous datasets (CMPIP6 and ERA5) covering different variables, spatio-temporal coverage and physical underpinnings. 

 

ClimaX aims to replace

 
 
 
Previous
Previous

A better way to study ocean currents

Next
Next

Detection of Bow Echoes in Kilometer-Scale Forecasts Using a Convolutional Neural Network